korrents

korrents · piece

Why are AI agents lying, cheating and coordinating?

Yoshua Bengio · 11 Sept 2026

Read the piece on yoshuabengio.org ↗

3 korrents from this piece ↓

Yoshua Bengio did not write this page.

Every claim below was made in this piece, quoted word for word and numbered in the order the piece makes them, so you can read it there rather than take our word for it. The sentence above each quote is our reading of the claim, not their wording. Each quote was checked against a stored copy of the page at build time; where the two differ, the quote is the fact.

  1. as AI capabilities keep growing, this kind of behavior could keep growing in severity too, unless we revisit the principles by which the most advanced models are trained.
  2. staying in operation, learning about the world and gaining control over it are stepping stones toward almost any other goal.
  3. So a more capable agent is likelier to cheat than a weaker one, because it can find the loopholes the weaker one cannot.

Keep reading

The other pieces are the ones whose claims are worded most like this one's. That measures language, not agreement: they may argue the opposite.